The exact llama.cpp (qwen/4090) and vLLM (hermes/3090) launch commands,
recon'd 2026-05-20, so the foxhop-states salt states can be written
accurately rather than guessed. Notes the convention (salt manages the
systemd unit; engine binaries + model artifacts stay manual on
/mnt/data as documented prereqs) and the live-hermes cutover constraint
(keep >=1 hermes online; qwen is expendable).
fox's call: capture the RAPL permission as a Makefile target run with
sudo, not an ad-hoc chmod. `sudo make rapl-access` on each GPU box
installs a udev rule (/etc/udev/rules.d/99-rapl-readable.rules) that
makes intel-rapl energy_uj world-readable on every powercap add event
(survives reboot), and applies chmod immediately so no reboot is
needed. Idempotent; reversible via `sudo make rapl-access-revoke`.
energy_uj is root-only by default (PLATYPUS side-channel mitigation,
CVE-2020-8694) — that's why bench/watt_bench.py's CpuSampler read null
CPU watts as fox during recon. After this target runs, watt_bench
reads CPU package energy directly (no --cpu-energy-cmd needed). GPU
watts via nvidia-smi never needed special perm.
Non-root guard + help entries verified; Makefile parses clean.
Toward fox's next goal: score the full serving stack on quality AND
cost — {qwen, hermes} × {llama.cpp, vLLM} × {3090, 4090} × {solo,
arborist}, measuring CG% + GPU watts + CPU watts + joules/answer per
cell.
watt_bench.py — adds CpuSampler (Intel RAPL package energy via
/sys/class/powercap/intel-rapl:*/energy_uj). RAPL exposes a cumulative
microjoule counter, so energy-over-window is an end-minus-start diff
(handles wrap) — more accurate than integrating instantaneous power.
Sums multi-package. energy_uj is root-only by default (PLATYPUS /
CVE-2020-8694), so it degrades to available=False when locked;
--cpu-energy-cmd 'sudo cat {path}' supplies a privileged reader when a
sudo rule exists. Each cell now reports gpu/cpu/total joules-per-
question + gpu joules-per-token; the report records cpu_rapl_available.
Verified: graceful degradation when locked; RAPL diff math (1->4 MJ uJ
= 3.0 J, exact).
benchmark-matrix.md — expands the cost section to the full 16-cell
(model × engine × GPU × arm) design, the per-cell metric set (quality +
GPU + CPU energy), the serving-stack inventory from 2026-05-20 recon
(4090=qwen/llama.cpp, 3090=hermes/vLLM — each box has one engine + one
model today), and the buildout gap (vLLM+qwen, llama.cpp+hermes, cross-
GPU models). Notes idle-floor asymmetry (hermes/3090 ~127W vs
qwen/4090 ~20W) as a real optimizer input.
Harness is ready; the serving-config buildout + RAPL perm grant are the
remaining (ops, fox-directed) prerequisites to run the full matrix.
Two deliverables for the cost/energy axis of the constraint optimizer.
docs/benchmark-matrix.md — shareable spec of the control experiment:
the question, fixture (386 office-holder Qs with corpus-vintage gold),
the 3-model × 3-framing × 2-arm matrix (18 cells), the verdict
vocabulary + two reads (accuracy vs grounding-fidelity), the
deterministic code judge + its Opus calibration, the results-so-far
table, and the NEW cost dimension (tokens / latency / GPU watts /
joules-per-answer measured per GPU tier). Self-contained — readable
cold by David.
bench/watt_bench.py — GPU wattage harness. Samples nvidia-smi
power.draw on the inference GPU while driving a small representative
subset, reports mean/peak watts, trapezoid-integrated joules,
joules-per-question, and joules-per-token. Tags the GPU
(--gpu-label 3090|4090) so the optimizer can compare hardware tiers.
Idle-baseline sampling separates load draw from idle. Does NOT grade
(energy is independent of correctness); saves answers + per-question
timing to JSONL for a later quality-per-joule pass via
score_with_code_judge.
Designed to run ON the GPU box (the orchestrator has no GPU; the
3090/4090 live on the inference boxes). Degrades gracefully when
nvidia-smi is absent (energy fields null) so it is testable anywhere.
Verified: PowerSampler graceful degradation + trapezoid integration
(synthetic 100->200->200W over 2s = 350 J, exact).
The headline cost finding the optimizer must weight: qwen-think
reasoning = 1300-3300 tokens/answer vs qwen-nothink ~50-100 (20-50x),
for a workload where arborist+qwen-nothink already lands 82% CG. The
energy numbers will quantify whether reasoning's premium is ever
justified — grounding-fidelity per joule, not per answer.
fox was right ("or you have defects still"): the arborist+qwen-think
empties were NOT an inherent reasoning limitation, they were
max_tokens set too low. Diagnosed with finish_reason + token usage on
a realistic large arborist-style context:
max_tokens=1024 → 4/4 empty, ALL finish_reason='length', all hit
exactly 1024 tokens
max_tokens=4096 → 0/4 empty, ALL finish_reason='stop', used
1339-3295 completion tokens
qwen-think spends 1300-3300 tokens on its (internal) reasoning trace
BEFORE emitting the tiny claim-lattice JSON. The arborist arm's budget
comes from DEFAULT_QUERY_POLICY["max_tokens"]=512 (correct for non-
reasoning single-line JSON), so reasoning refs hit finish='length'
mid-trace and return EMPTY. Fix: arb_policy["max_tokens"]=8192 for
reasoning refs (generous headroom over the 3295 observed ceiling).
This also CORRECTS a wrong claim in commit aa9d9c8's message: the
json-schema grammar does NOT suppress the reasoning trace. The
reasoning happens (it burns those 1300-3300 completion tokens); the
grammar only shapes the final emitted answer into schema-valid JSON.
So arborist+qwen-think is a genuinely distinct config, not a clone of
arborist+qwen-nothink. The earlier "grammar suppresses reasoning"
read was an artefact of toy-prompt diagnostics where reasoning fit
under 1024 tokens.
The empty-retry from aa9d9c8 stays as belt-and-suspenders for any
residual model-side empties, but with an adequate budget it should
rarely fire.
Cost note (fox's standing point, now quantified): reasoning = 1300-
3300 tokens/answer vs qwen-nothink's ~50-100 = 20-50x token cost. For
a workload where arborist+qwen-nothink already lands 82% CG, that cost
multiplier is the open question — not whether the cell is measurable
(it now is), but whether the marginal lift justifies 20-50x spend.
Two defects blocked the arborist+qwen-think cell, both diagnosed
2026-05-20:
Defect 1 — stop sequence truncates to empty. The claim_lattice path
sets stop=['\n\n'] (runaway guard tuned for single-line Hermes JSON).
A reasoning model's output trips it immediately → 100% empty answers
→ 100% ABSTAINED (measured on the first 23 items of the killed run).
Fix: MODELS gains an explicit flag; the arborist arm
clears claim_lattice_json_stop_sequences for reasoning refs. Direct
A/B confirmed: stop=['\n\n'] → ''; stop=None → valid JSON.
Defect 2 — intermittent empty completions. Even with the stop cleared,
qwen-think under json-schema grammar emits an empty completion ~1/3 of
calls (a llama.cpp reasoning+grammar artefact; qwen-nothink phase 3
had ~0 spurious empties). Fix: empty-output self-heal — reasoning refs
retry up to 3 attempts, burning the cached empty each retry. Never
fabricates: a still-empty answer after retries is recorded as empty.
6-item smoke: 0/6 (broken) → 5/6 valid JSON; residual ~1/6 are
questions that reliably break (4 consecutive empties), a documented
artefact.
Structural finding (to fold into Addendum 8): json-schema grammar
enforcement SUPPRESSES the reasoning trace — output is pure single-
line JSON, no <think> block. So arborist+qwen-think is structurally
≈ arborist+qwen-nothink; the thinking lever that moved the solo arm
is neutralized by the claim_lattice grammar. The re-run will confirm
empirically.
Non-reasoning refs (hermes, qwen-nothink) unchanged: single pass,
stop sequence intact.
Durable record of the #000057 sweep in the bench journal. Captures:
- The question: is Hermes-8B's confident present-day-officeholder
fabrication an 8B weakness, a framing artefact, or does retrieval
fix it? Crosses {hermes, qwen-nothink, qwen-think} × {plain,
source_relative, as_of_corpus} × {solo, arborist} on a 386-item
office-holder fixture with corpus-vintage gold.
- The judge methodology: Opus headless judge burned quota (79.5%
JUDGE_ERROR), replaced with the deterministic code judge
(bench/judge_code.py), calibrated against Opus's gradeable records
(CG agreement 13->47%, WRONG 56->89%, ABSTAINED 80->95%).
- Consolidated CG% scorecard, all arms on the identical final judge.
- Three findings:
1. Retrieval dominates — arb/qwen-nothink/plain 82% vs 7% solo;
no solo config approaches the retrieval arms.
2. Reasoning does NOT improve raw correctness — qwen-think/as_of
44% vs nothink 50%.
3. Reasoning's real cost is broken honest-abstention —
qwen-nothink/source_relative abstains 97% (clean); qwen-think
only 61%, reasoning itself into wrong parametric answers.
- Production recommendation: arborist + qwen-nothink, plain framing,
reasoning OFF (82% CG, ~0% abstain, 11% wrong-assert).
- Held cell noted: arborist+qwen-think running at write time, result
to be appended.
Bench %s are point-in-time measurements (not repo-derived counts),
so no AUTOCOUNT tags — consistent with addenda 1-7. test_doc_counts
3/3.
fox 2026-05-19: 'we don't need to redo anything'. The arborist+qwen-
nothink sweep would re-run qwen-nothink solo on 1158 records that
phase 1 already produced (control_sweep_2026-05-19T21-52-56Z.jsonl,
calibrated rescore on disk). Wasteful — qwen-nothink solo behaviour
doesn't depend on retrieval being on, so phase 1's numbers stand.
--skip-solo runs ONLY the arborist arm. Implementation: skip the
per-variant model loop in _process_item when the flag is set; the
arborist arm below it still runs if arborist_on. The spend banner
zeroes the solo-call count so the operator sees the actual LLM
budget for the arborist-only run.
Example: a full arborist+qwen-nothink sweep against the 386-item
fixture goes from 2316 LLM calls (1158 solo + 1158 arborist) to
1158 calls (arborist only). Halves wall-clock on single-worker
llama.cpp.
pytest 27/27, imports clean, --help shows the flag.
Reorder rule 3 (short-answer entity grounding) above rule 4 (NLI
contradiction) so positive lexical evidence cannot be overridden by
NLI clause-level noise. Surfaced by the 2026-05-19 arborist+qwen-
nothink smoke:
i=3 · who is the prime minister of Poland?
ans: 'Donald Tusk is listed as the Prime Minister of Poland.'
gold: ...lists Tusk + Marcinkiewicz + Belka + Kaczynski + Kopacz...
NLI contradiction p=0.892 (above 0.85 threshold)
NLI entailment p=0.744 (also high on the correct clause)
Tusk WAS PM in 2010 (served 2007-2014); answer is correct against
the corpus-vintage gold. The NLI contradiction signal came from
clause-level candidate selection picking a NON-Tusk PM the source
also mentions; entailment was high on the Tusk clause. Mixed signal
that the WRONG rule then over-confidently resolved.
The fix is a rule reorder, not a threshold change — the fast path's
positive-evidence combination (specifics-in-gold AND subject-in-gold)
is a strictly stronger signal than NLI's clause-level max
contradiction, so when it fires it should win. The combination
discriminates Poland-Tusk (Tusk ∈ gold, Poland ∈ gold → CG) from
Anthony-Albanese (Albanese ∉ gold → fast path declines → falls
through to UNGROUNDED-subject-in-gold → WRONG, unchanged).
Self-test 4/4 INSTRUMENT TRUSTWORTHY unchanged. pytest 27/27.
Poland-Tusk regression smoke: now CG via short_entity_grounded ✓.
No regression risk on the existing reconciliation cells:
- Iceland CG: short_entity_grounded was already winning (was rule
4, now rule 3 — same outcome, earlier exit)
- WWII-1812 WRONG: '1812' ∉ gold → fast path declines, NLI fires ✓
- Higgs-cafe FABRICATED: 'Higgs' ∉ gold → fast path declines ✓
- Anthony Albanese WRONG: 'Albanese' ∉ gold → fast path declines ✓
- Abstention phrases: rule 2 still fires first ✓
Two surgical fixes unblock 'arborist with synthesis LLM = Qwen-on-
llama.cpp' as a viable arm in the control sweep. Pre-existing
docstring said 'Arborist×Qwen needs proof-path guided_json+extra_body
surgery — coupled follow-up'; this is that follow-up.
Fix 1 — multi-engine structured-output extras
The runner / query JSON-mode paths previously sent only vLLM's
'guided_json' key for the claim_lattice schema. llama.cpp silently
drops it, leaving Qwen un-enforced (the parse-tolerant fallback did
all the work). Helper
claim_lattice_structured_output_extras() in arborist/qa/verify.py
now returns a dict carrying the schema under all three engine
conventions:
- guided_json (vLLM grammar-constrained sampling)
- json_schema (llama.cpp native shorthand)
- response_format (OpenAI-spec, honoured by llama.cpp and newer vLLM)
Each engine recognises its own key and silently drops the others.
Used at both inference call sites (runner.py:740, query.py:3324).
Hermes/vLLM path is unchanged — it picks up 'guided_json' and
ignores the other two.
Fix 2 — query() accepts user-supplied extra_body, merges with defaults
query() grew a keyword-only extra_body parameter (default None).
Per-model knobs (Qwen's {'chat_template_kwargs': {'enable_thinking':
False}} toggle, future template knobs) can flow from the caller to
the synthesis chat-completion call. Schema-enforcement extras are
added inside query() and merge under user keys — common case is
disjoint namespaces, but if a caller wants to override 'guided_json'
they can.
bench/control_sweep.py now passes MODELS[arborist_ref]['extra']
through to query() in the arborist branch, so --arborist-ref
qwen-nothink runs with reasoning disabled and --arborist-ref
qwen-think runs with reasoning enabled. Phase 1's arborist arm with
--arborist-ref=hermes is unaffected (MODELS['hermes']['extra'] is
None, merges to no-op).
Tests
+ 3 new in tests/test_verify_json.py covering helper default shape,
alternate-schema reuse, and query()'s new extra_body parameter
220 affected tests still green (verify / claim_lattice / judge /
runner suite)
pytest test_verify_json: 27/27
Next: small smoke run --arborist-ref qwen-nothink against 4-8 items
to confirm end-to-end before any full sweep. Phase 2 (qwen-think solo)
still running in background, unaffected — it doesn't touch the
arborist arm.
The Arborist arm runs answer_mode='claim_lattice' (per control_sweep.py
:179, control_ab.py:155) so its answers arrive as the JSON envelope
{"claims":[{"text":"...","evidence_ids":["E1"]},...]}.
_descaffold strips the [E1] evidence-pointer markup but the JSON
braces + key syntax remain. The verifier's strategy-2 (span) and
strategy-3 (proper-noun) extractors see brace noise instead of the
inner claim prose — every Arborist record degraded to UNGROUNDED.
The 2026-05-19T17-01-17Z sweep, re-graded with the freshly calibrated
judge (5a17f61), surfaced this: Arborist arm reported 0 CG across all
three variants in the live phase 1 output (the live run was pre-
calibration), and 29/120 CG (24%) under the calibrated rescore — clear
improvement just from theta_contra=0.85, but the JSON envelope was
still hobbling the verifier paths.
Fix: _unwrap_claim_lattice_json runs BEFORE all downstream rules.
Detection is conservative (three independent signals: starts-with-
brace AND "claims" key AND "text" key) so plain-prose answers
pass through unchanged. Multi-claim envelopes concatenate as discrete
sentences (extract_claim_spans treats each as its own span).
Malformed JSON falls back to the original answer — no silent
rewriting on broken input.
Smoke result on the Iceland Arborist case
ans: {"claims":[{"text":"The current president of Iceland is
Ólafur Ragnar Grímsson.","evidence_ids":["E1"]}]}
gold: {{Infobox Political post |post = President |body = Iceland
|incumbent = [[Ólafur Ragnar Grímsson]] ...}}
before: UNGROUNDED → FABRICATED (then WRONG after calibration)
after: short_entity_grounded → CORRECT_GROUNDED
pytest: 27/27 (added 7 unwrap-coverage tests covering single-claim
envelopes, multi-claim concatenation, plain-prose passthrough,
malformed-JSON tolerance, unrelated-JSON passthrough, and the
end-to-end Arborist-envelope CG flow). Self-test 4/4 unchanged.
Re-rescores of 17:01 sweep + phase 1 sweep run after this commit
to measure final Arborist scorecard improvement.
Opus is trusted; this commit closes the systematic gaps surfaced by
B's rescore on the 17:01 sweep (2289 records, of which 468 had real
Opus verdicts). Off-diagonal cells in descending size and the
root-cause fix for each:
cell n=114 code:WRONG · opus:CORRECT_GROUNDED
Root cause: NLI fires contradiction p in [0.5, 0.75] on factual
answers like 'Ólafur Ragnar Grímsson is president of Iceland'
against wikitext-shaped infobox gold — clause-level candidate
selection picks up co-mentioned earlier office-holders, reads
temporal-frame mismatch as contradiction. The TRUE contradiction
signal (WW2 1812 self-test fixture) measures p=0.985 — clean
margin above noise.
Fix: raise theta_contra 0.5 → 0.85 (code-judge override of the
NLI manifest's 0.5 contradiction_veto).
cell n=60 code:FABRICATED · opus:WRONG
Root cause: 'Anthony Albanese' answer vs Julia Gillard gold gets
FABRICATED (specifics-not-in-gold) but Opus correctly distinguishes
WRONG (source contradicts by naming someone else) from
FABRICATED (source silent on topic).
Fix: when verifier UNGROUNDED + specifics not in gold AND the
question's subject anchor IS in gold, demote FABRICATED → WRONG.
Subject anchor uses proper-noun-shaped terms from the question
(Iceland / Australia / Higgs) — not the last-content-token
heuristic, which mis-fires on coincidental matches like 'cafe'
appearing in a 'gold does not mention any cafe' denial.
cell n=18 code:WRONG · opus:ABSTAINED
Root cause: abstention patterns missed Hermes's most common
refusal phrasings — 'I do not have accurate information', 'I do
not have access to a reference knowledge base', 'I lack access
to'. Original patterns required determine/know/tell verbs right
after 'do not'.
Fix: three new patterns for the 'do not have / lack ...
information / access / knowledge' family.
cell n=13 code:ABSTAINED · opus:CORRECT_GROUNDED
Root cause: verifier's strategy-2 needs prose shape; terse-name
answers ('Pratibha Patil', 'Jalal Talabani') fall to
UNGROUNDED-no-specifics → ABSTAINED, missing valid CG.
Fix: short-answer entity-grounding fast path. When answer is
short (≤15 tokens) AND every specific asserted is present in
gold (no unsourced) AND at least one specific WAS asserted AND
the question's subject anchor is in gold → CG. Guards against
'wrong topic, right name' false-positives via the subject check.
Structural reorder: NLI contradiction now runs AFTER the
abstention check and short-answer fast path (instead of preempting
the verifier), so the verifier's STRICT/HYBRID positive signal
isn't overridden by NLI noise. NLI still leads the path on truly
unbounded answers — verifier UNGROUNDED + NLI ≥ 0.85 contradiction
keeps the WRONG label.
Self-test 4/4 INSTRUMENT TRUSTWORTHY. pytest contract 18/18.
v2 rescore on the same 17:01 sweep runs in the background to
measure agreement-matrix improvement empirically.
Added bench/analyze_judge_disagreement.py — the harness that drove
this calibration (reads B's rescore JSONL, bucketises off-diagonal
cells, dumps configurable samples per cell with question / answer /
gold / both rationales). Reusable for the next calibration round.
Reads an existing sweep JSONL (Opus verdicts already recorded), re-fetches
gold per record via _gold(), runs the code judge on (question_asked,
answer, gold), and emits:
- markdown scorecard: agreement matrix (code × original judge), per-arm
/ per-model / per-variant code-judge tallies, and a residue table of
the JUDGE_ERROR records (the natural input to a later LLM-batch
needle-haystack pass — Opus or Grok);
- JSONL with one row per sweep record (code_verdict + code_rationale
+ code_decision), joinable on (i, arm, model, variant) to the
source sweep.
Zero LLM calls. Reads sweep JSONL + shards read-only. Pairs with the
new --judge switch (a2e9b49): the switch decides what NEW data uses;
this script decides what the ALREADY-COLLECTED data looks like under
the deterministic judge.
Usage (parameter default matches control_sweep.py default fixture):
python -m bench.score_with_code_judge --in <sweep>.jsonl
Currently running against control_sweep_2026-05-19T17-01-17Z.jsonl
(the 2289-record sweep that ran on the prior huge-N pass before the
Opus quota burned out). Output will land at
bench/qa_results/control_sweep_2026-05-19T17-01-17Z_code_judge.{md,jsonl}.
Wire bench/judge_code.py into the sweep harnesses as the default judge.
Both control_sweep.py and control_ab.py grow a --judge {code,opus} CLI
arg; both share the same Verdict shape so the dispatch is a pointer
assignment + threading the judge_fn through _process_item.
Behaviour:
- DEFAULT = code: zero LLM, zero quota, deterministic. Self-test gate
is the code judge's 4-fixture contract.
- --judge opus: original gated Opus path; needs ARBORIST_JUDGE_ENABLE=1
set per 1cabfe6's fail-closed gate, otherwise every record returns
JUDGE_ERROR with rationale 'disabled — set ARBORIST_JUDGE_ENABLE=1'
and the sweep records that label honestly.
Reporting:
- Header line now records which judge ran ('Judge = code (...)' or
'Judge = opus (...)') so partial-reports & resumes don't lie about
provenance.
- Spend banner shows '0 LLM calls' for the code path so the no-burn
property is visible in the operator output.
Test surface: pytest sweep across tests/ still 136/136 (no regressions);
new --judge flag visible in --help on both harnesses.
Next: bench/score_with_code_judge.py to re-grade existing sweep JSONLs
(written under the gated-Opus run) with the code judge; agreement
matrix surfaces residue size for the eventual LLM-batch needle-haystack.
bench/judge_code.py — drop-in alternative to bench/judge.py with the
same Verdict shape & closed verdict vocabulary (CG/W/F/A/JE) but zero
quota cost: composes verifier + NLI + abstention + specificity into a
fixed-order pipeline. fox 2026-05-19: 'data first, judging later' —
this is the data-collection arm; LLM-based judging (Opus batched
needle-haystack, or Grok credit-card) is a separate downstream
concern that operates on the residue this judge cannot classify
deterministically.
Pipeline (first hit decides):
1. empty / no-gold guards
2. explicit abstention phrases (lexical regex)
3. NLI contradiction (arborist.qa.nli.shadow_check) — strongest
signal: gold contradicts the claim → WRONG
4. lexical verifier (arborist.qa.verify.verify_quotes) →
STRICT → CORRECT_GROUNDED
HYBRID + NLI entail >= 0.55 → CORRECT_GROUNDED
UNGROUNDED + specifics-not-in-gold → FABRICATED
UNGROUNDED + no specifics → ABSTAINED
HYBRID without NLI corroboration → JUDGE_ERROR (residue
for an LLM judge)
Threshold note: _CODE_JUDGE_THETA_ENTAIL_CORROBORATE=0.55 is distinct
from the NLI manifest's entailment_block_veto=0.9. The manifest's
threshold is calibrated for OVERRIDING a STRICT lexical signal with
negative evidence — high bar. The corroboration use here is the
opposite direction: additive positive evidence on an already-positive
anchor — moderate bar appropriate. Self-test case 1 measures NLI
entail=0.769 (clearly entailed, clear margin above 0.55).
Specificity for FABRICATED layers three scanners:
- verifier's multi-word proper-noun extractor (Higgs Boson, ...)
- local single-word capitalised-token scanner (Napoleon, Mars, ...)
deliberately separate because the verifier's gate is conservative
by design (multi-word only)
- numerics (years, dates, large counts, money)
Self-test: same 4 fixtures as bench/judge.py:self_test() so the two
instruments can be cross-checked when fox re-fires the Opus judge on
the residue later. Result: 4/4 INSTRUMENT TRUSTWORTHY.
tests/test_judge_code.py — pulls the contract into make test
(18 cases): module identifiers pinned, dataclass shape parity,
empty / no-gold guards, parametrised abstention phrases, specificity
layer behaviour, the canonical 4-case self-test, batch helper, and
graceful NLI-unavailable degradation. 18/18 pass.
Pre-existing known limitation, documented in the docstring: terse
correct answers ('In 1945.' against gold containing '1945') route to
ABSTAINED because the verifier's span extractor needs prose shape;
NLI sees no clause-level overlap at very short claims. The conservative
ABSTAINED label is correct deferral; tuning this is a calibration
question for real bench data, not the instrument's contract.
No callers touched yet — control_sweep.py & control_ab.py still
import the disabled Opus judge. Wiring this in is a separate ticket
move per fox's data-first sequencing.
2026-05-19: huge-N #000057 control sweep (f63b00d → 9dc02e4) burned
our Opus quota. Disable judge.py by default so a stray re-run can't
re-burn — every call short-circuits to JUDGE_ERROR with rationale
'disabled — set ARBORIST_JUDGE_ENABLE=1 ...' and zero subprocess
spawn (0ms in the disabled path, smoke-tested).
Why a gate, not a model swap:
- judge.py uses Opus deliberately as EXTERNAL SOTA outside both arms;
swapping the judge to Hermes/Qwen would corrupt the experiment
(Hermes is itself an arm under test). The hygiene comment at
judge.py:26-29 already names same-family-judging as the live threat
to validity at Opus level; downgrading further changes the science.
- Gating instead preserves the science when fox re-enables, and gives
us the data-first workflow he asked for: deterministic tool
pre-filters (verifier / NLI / recall@k) up front, judge only on
residue worth Opus tokens, with explicit go.
Behaviour:
- control_sweep.py + control_ab.py already treat JUDGE_ERROR
non-fatally (counted as JE in _bucket); disabled runs degrade to
100% JE in the tally and surface the disable reason in rationale —
the loudest possible 'judge did not run here' signal.
- Re-enable per-run: ARBORIST_JUDGE_ENABLE=1 python -m bench.control_sweep ...
- self_test() will report 4× JUDGE_ERROR when gated — intentional;
if the instrument is off, the self-test must NOT silently pass.
Smoke-test (without flag): label='JUDGE_ERROR' rationale='disabled — ...'
dt=0.0ms · no claude subprocess spawned.
Cross-referenced from CLAUDE.md '## Live endpoints' /
'Budget discipline' subsection added in 2365bd1.
2026-05-19: huge-N #000057 control-arm sweep (f63b00d → 9dc02e4) burned
our Opus quota. Pin the policy into repo CLAUDE.md so future agents
don't reach for Opus on autopilot:
- Sweep + bench-qa default endpoints: Hermes-3-8B + Qwen via uncloseai
- Opus drops out of control-arm grids until fox explicitly re-adds it
- Existing bench/results/ data stays valid; just no new Opus expansion
- Translation still uses local pinned opus-mt (sandwich MT edges), not
Hermes/Qwen 8B
- Grok rejoin path noted as an HTML comment (paid spend, fox-gated)
Adds two subsections under '## Live endpoints':
- 'Budget discipline — Hermes / Qwen first, Opus deferred'
- 'Per-call model selection (general agent work, not sweep arms)'
fox: 'not 11 hours it shouldn't take that long'. Probed endpoints —
Qwen-27B absorbs 8 concurrent with 0 errors (0.5->2.1s); the
bottleneck is the serialized claude -p Opus judge, so more workers
≈ near-linear speedup. control_sweep.py gains --resume PATH: appends
to an existing JSONL, skips items already COMPLETE (full
models×variants for solo + variants for arborist if i<=arborist-n),
re-runs partial items; _aggregate now dedupes (i,arm,model,variant)
last-wins so a killed-mid-unit restart never double-counts, and
_load_recs tolerates a truncated trailing line from the kill.
Makefile control-sweep gains CONTROL_SWEEP_WORKERS / _RESUME / _ARB_N
so make stays the interface. 6-worker run killed cleanly (specific
pids, no pkill), relaunched resume @ 12 workers — 12 done items
preserved, 374 to run, ~5h -> ~2-2.5h.
fox: 'make the n huge huge, check in every ~7 turns'.
- stale fixture re-mined to the FULL pool: 386 questions (180 pres /
119 PM / 34 premier / 20 gov-gen / 17 chancellor / 15 CM / 1 FM),
deterministic.
- control_sweep.py rebuilt: ThreadPoolExecutor work-unit-per-item
(per-unit qa_db -> zero SQLite write contention; shards read-only
safe under concurrent readers), incremental flushed JSONL, and a
--report-only PATH mode that aggregates a PARTIAL file with ZERO
LLM/judge spend (the interim check-in path).
- huge N goes on the CONTROL (solo x3 models x3 framings) where the
open statistical question lives; Arborist A/B = fixed --arborist-n
(default 40), not re-measured 386x (power belongs on the control,
and query() over ~40GB shards is the heaviest call).
- bench-maxing doctrine applied: independent hermetic judge calls
fanned out, 'serial-by-caution is halting in disguise'.
N=3 (prior run) already shows the coherent, review-relevant story:
plain -> all 3 models confidently assert (0% abstain); source_relative
-> abstain jumps (hermes 33 / qwen-think 67 / qwen-nothink 100%);
as_of_corpus -> ~100% correct all models. The gap is largely
framing+snapshot, not a universal capability deficit.
fox ruling: 'we both do not know which framing is right, measure all
benchmarks and bring results forward for review'; 'we have qwen with
and without reasoning to use'.
bench/control_sweep.py: sweeps the CONTROL (solo) arm across
{Hermes-3-8B, Qwen3.6-27B reasoning, Qwen3.6-27B no-reasoning} ×
{plain, source-relative, as-of-corpus-era} questions, judged vs the
fixed corpus-vintage gold; Arborist-Hermes treatment reference run
alongside at the same N. Presents the SAME judge verdicts under both
the accuracy framing (the naive read fox flagged as unfair-as-truth,
shown for contrast) and the grounding-fidelity framing (the
defensible read), plus an explicit note on the faithfulness-ablation
framing + Arborist×Qwen — both deliberately NOT run (proof-path
surgery / different instrument, not a relabelling).
Answers fox's open question empirically: does a 27B *reasoning*
control honestly ABSTAIN where the 8B fabricates (gap = weak-small-
model artefact) or still confidently assert post-corpus (gap = real,
scale-independent)? as_of_corpus separates 'can't recall the era'
from 'won't constrain to a source'.
Qwen toggles probed live: reasoning answer in message.content (CoT in
separate reasoning_content, not surfaced — logged limitation);
no-reasoning via chat_template_kwargs{enable_thinking:false}. Gate =
in-script judge self_test (aborts on fail) — stronger than a make
edge, no double self-test spend.
Opt-in third NLI inference backend (ARBORIST_NLI_BACKEND=tinygrad) in
qa/nli/shadow.py, parallel to torch/onnx-int8, behind the #000049
cage: shadow-only, never an audit_mode input, never auto-preempts the
proven path (guarded so it cannot regress torch/onnx). Loads the ONNX
export through tinygrad's frontend wrapped to the existing
model(**enc).logits contract so _nli_batch is byte-unchanged.
bench/scripts/nli_backend_ab.py + make bench-nli-backends: deterministic
A/B, gate is numerical agreement with the torch reference first,
latency second (a divergent engine = a different shadow signal = a
different nli_policy_hash). Instrument is honest — reports a requested
backend as unavailable rather than relabelling a fallback's numbers.
First CPU-smoke run already quantified that the deployed §7 #22 int8
export diverges Δmax≈0.42 from torch — the immunity property made
measurable, not a defect. Real tinygrad numbers pending a producer-box
run (tinygrad not an arborist dep; frontend op-coverage for the large
MNLI checkpoints unverified by design).
docs/onnx-vendor-capture-immunity.md: why the model-in-proof-path cage
makes the inference engine an interchangeable sidecar, never a trust
dependency — public-domain positioning capital. Indexed in CLAUDE.md.
Full suite 2498 passed (identical to baseline); 24/24 NLI tests green.
control_ab.py called client.chat_completion(messages) but model is a
keyword-only REQUIRED arg → TypeError every call → solo answer was
always '[solo-error: TypeError]', which the hermetic judge correctly
classified ABSTAINED. Every prior 'Hermes-solo abstained' smoke
(incl. the n=2 the ticket cites as a case against the treatment) was
a broken-control-arm artefact, not a measurement. Fix: pass
model=a.model (symmetry with the Arborist arm). Error token now
carries the exception message so the next failure self-describes.
Also adds the #000057 'stale' fabrication-bait mining class
(time-varying executive offices; corpus = ~2010-2011 Wikipedia
VERIFIED from the artifact, NOT the 2003 dump CLAUDE.md names —
flagged) and restores the numeral fixture deterministically.
b573c59's accent-fold work removed 210 lines from
qa_questions_numeral_map.json (40 -> 5 entries) — collateral of an
earlier `mine_questions.py --cls numeral --limit 5` regression-check
that got committed. The control-ab run scored n=5 (not the --n 12
passed) because the fixture itself only had 5 rows. Re-mined at the
default --limit 40; deterministic mine yields a byte-identical
restoration of the a3ac653 original. Caught by reading the artifact,
not assuming.
bench/control_ab.py + `make control-ab` (gated on judge-self-test
as a make dependency — instrument gate cannot be skipped). Same
model both arms; gold = target-article text by target_root; Arborist
[E…] scaffolding stripped (blinding — format can't betray the arm);
Arborist UNGROUNDED credited as honest abstention; hermetic Opus
judge; deterministic aggregate; self-auditing JSONL; threats-to-
validity printed in the report.
N=2 smoke: clean end-to-end, 0 JUDGE_ERROR — and already surfaced a
case AGAINST the treatment (solo correctly ABSTAINED; Arborist
HYBRID-WRONG). The instrument can falsify the Arborist value claim;
that is the point. n=2 proves nothing (report says so) — verdict
needs a real N.
fox ruled judge = Opus via `claude -p`. bench/judge.py:
hermetic (`env -u CLAUDECODE claude -p`, fresh process, context =
only (Q, answer, gold) — no arm label, no Arborist context, no
session), blinded-by-caller, reference-grounded against the fixed
gold (ignore parametric knowledge), structured via FINAL_VERDICT=
sentinel parsed LAST-match.
Instrument-before-experiment gate worked: first cut parsed
first-match over the model's chain-of-thought → 0/3 self-test. The
judge REASONED correctly; the parser was the defect (+ two bad test
fixtures, my error). Hardened (sentinel contract + fixed fixtures),
re-verified: `make judge-self-test` = 4/4 on known-verdict triples
via real claude -p. The make target is the precondition gate; no
control run trusts the judge until it passes.
Threat to validity recorded, not hidden: same model family judging;
mitigated (blind + no-stake + reference-grounded) not eliminated —
different-family SOTA cross-check is the only full removal.
Next: bench/control_ab.py + `make control-ab` (Hermes-solo vs
Arborist, gold=target-article text, blinded, judged) — NOT yet
built; no broken make target shipped for it.
fox: should the judge get the same docs, or the 8 retrieved/cited?
Neither. Per-arm retrieved docs (a) are undefined for Hermes-solo
(no retrieval → breaks A/B symmetry), (b) re-import circularity
(judging the treatment vs its own retrieval), (c) measure citation
self-consistency (Arborist's verifier already does that), not
truth. Judge's reference = the mined question's fixed ground-truth
target/gold, identical for both arms, independent of retrieval —
the ground-truth-carrying fixture supplies exactly this. Recorded a
distinct, non-headline faithfulness ablation (judge Arborist vs its
own context) so it is never conflated with the control delta.
Index row unchanged (skim surface; ticket body is the design log).
I over-applied a proof-path rule. "No LLM-as-judge" guards Arborist's
*verifier* (an LLM in the proof path = the circularity Arborist
kills; #000049's poor reliability is a proof-path-model constraint).
The control experiment's judge is external science — outside BOTH
arms, scoring post-hoc, touching neither system's internals — so it
cannot break the experiment. fox: use a strong SOTA model
(stronger than both systems under test), blinded + reference-
grounded + transcript-logged. This lifts the wrongly-imposed
atomic-fact-only limit (judge handles essay grounding; #000049's
exclusion was right for the proof path, wrong for external science);
deterministic gold-key demoted to a cheap atomic cross-check.
Still fox calls: fixture population, which SOTA judge (API
cost/egress), spend. Scaffold; no build past gate-1.
The review's strongest point (§4.1) + fox: every measurement so far
is WITHIN Arborist — no control arm. Folded into #000057 (not a new
ticket): control = Hermes-solo; judge external/deterministic/
paraphrase-tolerant via a mined gold-fact answer key (resolves
circularity + no-LLM-judge + paraphrase at once); headline metric =
fabrication-vs-honest-abstention delta, not raw accuracy (common-
knowledge fixture would falsely show null — parametric memory).
Smallest-proof reorder: v1 = single-shot Hermes-solo vs Arborist on
a fabrication-bait atomic-fact fixture; recursive-drift loop is v2,
gated on v1. Atomic-fact only; essay-grounding = #000049's unsolved
NLI, OUT. Still fox calls: fixture population, gold-key shapes,
spend. Scaffold; no build past gate-1.
The 2026-05-19 GPT-5.5 Hamming-framed review's ONE arborist-scoped,
ticket-worthy nugget: prove the Merkle-Providence-Reverse-RAG
whitepaper's headline claim (untracked evidence loss -> unbounded
recursive drift; witness-preserving state bounds it). Scaffold only,
awaiting fox go/no-go on scope.
Discipline encoded from the 2026-05-18 precedent (a grand target is
unfalsifiable until the instrument can resolve it — four hypotheses
died, only the deterministic mined-recall instrument broke it):
instrument-before-experiment, ONE task not eight, minimal ON/OFF
A/B, non-claims pinned (necessary substrate, NOT AGI).
Everything else in the review (rename, corpus hierarchy, IQ/talent,
ToE/Riemann/identity/geopolitics) deliberately NOT ticketed —
narrative/positioning, not arborist engineering; don't-proliferate.
Exactly one ticket. Next ID 000057 -> 000058 (same commit).
The review's central, correct finding: _title_query_tokens is
hot-path (every query AND title) and its fold set (hyphen #000007 +
numeral/accent/honorific/brit) changes which documents retrieve, but
that normalization's version was bound nowhere → a replay cannot
identify which token-normalization produced an old providence
record's sources. Same provenance class as the #000001 keyword gap.
Severity is honest: replay-provenance gap, NOT cache corruption —
different folds → different sources → different context_root →
different cache_key, so no false answer-cache aliasing or false
STRICT. Verifier/proof path unchanged.
Fix follows the repo's OWN #000001 §5/§6 decision (bind retrieval
transforms into the run-DAG RetrievalPlan/retrieval_plan_hash, NOT
governance_policy_hash). The review suggested governance "Option A";
repo precedent is run-DAG binding (same status as retrieval_keywords
and #000056 MT-engine identity) — the discrepancy is surfaced for
fox as an explicit call, not silently overridden.
- RetrievalPlan.title_token_policy (empty default → omitted from
canonical() → every prior retrieval_plan_hash byte-identical; the
§5 zero-churn discipline, same as the #000056 MT fields).
- _TITLE_TOKEN_POLICY single source of truth in query.py, bound at
the plan construction site; bump on any fold change.
- Plus the review's edge cases: Roman-substring-in-word not folded,
out-of-range not folded, Unicode Roman explicitly unsupported,
hyphen∘numeral composition. Full suite 2498, 0 regressions.
Declined (not engineering, per don't-proliferate): the review's
SelfModel/MemoryRoot/5S-5T-5F/capital-ledger ceremony — the ticket
design log is the single source of truth; scope recorded there.
Two more MEASURED fold-search wins on mined ground-truth fixtures
(deterministic recall, no LLM), both lifting at @1/@3/@8 (not
coarse-k artifacts):
honorific (Mt/St/Dr <-> Mount/Saint/Doctor): recall@1 45% -> 75%
(+30pp), @8 62% -> 85%, misses 15 -> 6
brit (British <-> American spelling): recall@1 50% -> 70%
(+20pp), @8 70% -> 85%, misses 12 -> 6
Both _*_fold_variants are additive+symmetric, strict closed sets
(no English-word collision), no-op outside their class (verified
independent: brit no-ops on honorific titles & vice versa), unioned
into _title_query_tokens beside hyphen(#000007)/numeral/accent.
Full suite 2488 passed, 0 regressions (hot-path); real-path tests
(FakeSource->ingest->query()->real _Hit).
Fold-search FINAL across the survey backlog, ranked by MEASURED @1
headroom (not prevalence — the instrument's job):
SHIPPED: numeral (a3ac653) accent (b573c59) honorific brit (here)
NO-BUILD: hyphen — existing #000007 already delivers 90%@1
(the measure-the-unmeasured-thing check pays off)
NO-BUILD: amp — 82%@1 with no fold (prevalence-overranked;
instrument killed it cheaply, like digit-ordinal pre-build)
Net: 4 deterministic retrieval wins + a reusable mined-recall
instrument + the discipline codified in CLAUDE.md, from a goal that
4 prior hypotheses died on because the bench couldn't measure them.
Second MEASURED fold-search win, and the instrument correcting my own
premature call. accent-fold ON vs OFF on the mined accent fixture:
recall@1 55% -> 85% (+30pp), rank-1 22/40 -> 34/40. recall@8 was
flat (95->98) — a too-lenient k nearly got a real lever wrongly
reverted; @1/@3 is the resolution that drives primary-source
selection. _accent_fold_variants: ASCII-fold then re-tokenise so a
diacritic title ("Béla Bartók", which _TITLE_TOKEN_RE otherwise
fragments to junk) matches the ASCII form a user types. Additive+
symmetric, no-op on pure-ASCII (zero effect on non-accent
queries/titles), mirrors _hyphen_fold_variants (#000007).
Also fixes a defect I shipped in a3ac653: an orphaned duplicate
body left as dead code after `return base` in _title_query_tokens
(unreachable — numeral-fold behaviour/measurement were valid — but
cruft; removed).
Fold-search factory, fanned out across the full survey backlog
(deterministic recall, no LLM, parallel — serial-by-caution was
halting in disguise):
- recall_at_k.py: returns rank -> recall@1/@3/@k from one retrieval
(verified offline). A coarse k hides rank-only lifts.
- mine_questions.py: numeral/accent/hyphen/honorific/amp/brit
ground-truth classes; fixtures committed.
- Measured @1 headroom verdicts: accent SHIP (this commit);
honorific 45% / brit 50% = real headroom (build next); hyphen
90% = existing #000007 already delivers, NOTHING to build (the
measure-the-unmeasured-thing check pays off); amp 82% = no fold
needed (prevalence-overranked, instrument kills it cheaply).
CLAUDE.md bench-maxing: two measured lessons codified — report
recall@1/@3/@k (a lenient k hides rank lifts; prevalence != miss-
rate), and fan out independent measurements (serial-by-caution is
halting). Full suite 2488 passed, 0 regressions (accent-fold is
hot-path in _title_query_tokens); real-path test (FakeSource->
ingest->query()->real _Hit).
Compressed headline: when a lever's failure class is ≤3-5/75 q the
curated n=3 audit_mode bench can't resolve it (four 2026-05-18
hypotheses died there). Fix the instrument: mine ground-truth-
carrying questions (bench/mine_questions.py) graded by deterministic
retrieval recall@k via `query --dry-run` (bench/recall_at_k.py) —
no LLM/verifier/n=3-noise/5pp-floor, ±1-q resolution. Caveats codified:
measures retrieval surfacing (necessary-not-sufficient for STRICT);
complements never replaces the curated adversarial trap gate.
The first MEASURED, above-noise retrieval win this thread. The 75-q
n=3 audit_mode bench couldn't resolve any single lever (every failure
class <=3-5 q, sub the 5pp floor — four hypotheses died there). Fix
the instrument, not just the lever:
- bench/mine_questions.py + bench/recall_at_k.py: mine questions from
corpus titles (ground-truth target known by construction), grade by
deterministic retrieval recall@k via `query --dry-run` — no LLM, no
verifier, no n=3 noise, scalable to the 22K-deep numeral pool. The
curated qa_questions.txt stays the separate verifier-honesty/trap
gate; mined fixtures measure the answerable long tail per class.
- _numeral_fold_variants in query.py: ordinal-word ("Alexander the
second") <-> multi-char Roman ("Alexander II"), additive+symmetric,
unioned into _title_query_tokens exactly like _hyphen_fold_variants
(#000007). Strict 2..40 Roman set → no English-word collision;
single-char Romans (I/V/X) intentionally out of scope (universal
len>1 token filter — stated before building, ~4 of 10 residual
misses).
Measured on the mined numeral fixture: recall@8 22/40 (55%) -> 30/40
(75%), +20pp; 20 hits now rank-1. Discipline applied end to end:
measured-first, mirrored precedent, full-suite regression run (2482
passed, 0 regressions — numeral-fold is hot-path in
_title_query_tokens), real-path test (FakeSource->ingest->query()->
real _Hit, not a hand-built object), measured-after on a noise-free
instrument. The ~6 multi-char residual misses are a different
downstream cause the instrument now exposes for future iteration.
The fr-no-mask sweep (2026-05-18) corrected an over-pessimistic read:
French was reported 47% but that was the failed-mask artifact; honest
fr no-mask is 61%. §9 + index now carry the firmed n=1 table (EN 85%
· es-nomask 71% · es+mask 65% · fr-nomask 61% · fr+mask 47%), the
comparator-corrected verdict (sandwich is a ~0%→61-71% net win vs the
real no-support baseline; the −14/−24pp is vs native English, not a
regression), and the entity-mask lever's confirmed net-negative
status (default-OFF). Doc-only; no code change.
Three workstreams, full suite 2482 passed, experimental paths default-OFF.
#000055 — Windows quickstart without make
tasks.py (pure-stdlib runner) + make.bat shim + .gitattributes;
README Windows section rewritten. Quickstart needs only Python
3.10+ (no make/bzip2/curl/bash). Mirrors the Makefile quickstart
subset; drift-pinned by tests/test_tasks_runner.py.
#000001 §7 Phase 0 — deterministic cross-language guard
arborist/qa/crosslang.py: non-English signal (¿/¡/non-ASCII) + an
es function-word stoppack. Fail-closed to UNGROUNDED before
retrieval/LLM (mirrors the quantifier reject-DAG) when no content
token survives, else strips es stopwords from the retrieval query
only. English path byte-identical by construction. Default OFF
(crosslang_guard_enabled). Measured: the anarcocapitalismo field
case 10.4s -> 1.6s.
#000056 — Operation Sandwich (cross-language grounding)
arborist/qa/mt/: opus-mt es/fr/ru<->en, lazy per-pair memoised
singleton (fixes the 88%-engine-error concurrency defect),
manifest-pinned, [mt] extra; entity_mask wrapper. Sandwich =
translate query in (retrieval + LLM prompt) -> English answer ->
UNTOUCHED verifier grounds English-vs-English -> translate the
verified answer out as display-only (banner-labelled, zero
grounding). question_hash + verifier_policy_hash invariant; MT
engine identity binds into RetrievalPlan, not governance. CLI
--crosslang-translate / make XLANG_MT=1. Default OFF; entity_mask
default OFF (measured net-negative at bench scale). Fan-out bench
(bench/*.py): Spanish ~0% -> 71% grounded vs the real no-support
baseline; the round-trip predictor was tried and refuted; the
entity-mask lever failed at scale (corpus-title anchoring untried).
CLAUDE.md: cross-language bright-line convention + module map.
Pre-existing modified diagram files are intentionally excluded.
Previously linked to the bare domain, which serves a marketing page. The actual OpenAI-compatible endpoint is /v1; /v1/models is the clickable verification (returns the served model card on the live deployment).
Shard capacity convention is documented in arborist/search/fts5.py:113 ("~50ms cold per token on a 10GB shard"). The earlier "~2 GB for a Wikipedia-sized corpus" line in both pagers was a fabricated figure off by 20×. Replaced with the real numbers: per-shard ~10 GB design target, live deployment of four shards totalling ~38 GB and holding 3.5 M documents / 6.2 M chunks.
gitlab-runner rotates between build dirs (builds/RUNNER_ID/0, /1, /3, ...). The cached .venv/ embeds the absolute build-dir path into __editable__.arborist-0.0.1.pth + the dist-info RECORD via pip's editable install. When the cache is restored into a different build dir, pip's implicit uninstall-then-reinstall step fails with `OSError: No such file or directory` looking for files at the old build dir.
Fix: shared .warm-venv-setup hidden job referenced from every real job's before_script. It wipes arborist-*.dist-info, the __editable__*.pth marker, the finder, and the bin/arborist entry script before pip install — so pip does a fresh install at the current build dir. Cache stays warm (deps don't re-download), only the editable-install bookkeeping is rebuilt (~1s per job).
DRY'd via `!reference [.warm-venv-setup, before_script]` so test / bench-suite / substrate-score / test-crawler all share the same setup; each job adds only its own install line ('.[dev]' vs '.[dev,crawler]').
Old drafts read as internal substrate notes. Rewrites lead with what the system does for a consumer or evaluator and what it costs to run, with no references to internal tickets, table names, schema-version strings, governance hash dimensions, or per-record audit-mode tokens. Appendix diagrams updated in lockstep: friendly labels ("grounded / partly grounded / not grounded") replace the schema-column trichotomy, layer names paraphrased away from SURFACE/CORE/PROVIDENCE.
1-pager (docs/_source/arborist-one-pager.rst, 1 page) for AI-literate readers: the trichotomy, the 8-dim cache key, CTI synthetic-elision-impossible, soft-channel separation, real-traffic bench numbers (mis-cite 100% @ 0% FP, warrant 92/92, quote 0.54 STRICT-rate).
2-pager (docs/_source/arborist-two-pager.rst, 3 pages = 2 body + 1 appendix) for technical reviewers: letterhead, Permacomputer Preamble license box, six numbered sections, plus appendix figures (pager-arch-stack 3-layer architecture, pager-verifier-flow question→pointer→verifier→trichotomy).
Both pages live under docs/_source/ so the same RST renders into the Sphinx readthedocs site (toctree caption "Summary pages" added to docs/_source/index.rst) AND into standalone PDFs via rst2pdf (docs/pager.style, lazy install into .venv).
Makefile targets: docs-one-pager, docs-two-pager, docs-pagers, docs-pagers-clean. Diagrams render through the existing DOT pipeline.
Synthesizes the bench-maxing work across:
#000049 NLI recombination veto (bart-large-mnli/k=12/margin/θ=0.999
→ 48% real-haystack recall at 0/808 STRICT FP — partial closure,
Phase-3 semantic candidate selector for full closure)
#000052 §3.1 diagnose_coherence (lexical sidecar, advisory-only,
1.1% real-STRICT FP after round-2 patch)
#000052 §3.2 relevance reranker (bge-reranker-large + cleaned +
θ ≤ -2.42 → 100% mis-cite / 55% deflection / 0% STRICT FP —
motivating Zionist-shape failure fully covered)
The three are architecturally orthogonal (§3.2.2 step 3C verified:
combining lexical sidecars with the relevance reranker gives no
lift; each owns its own failure-shape slice). Three structurally
distinct demote-only signals layered on the binary verifier.
Three runtime-promotion decisions for fox+dav1d:
- §3.1: keep advisory or wire policy hook? (probably advisory)
- §3.2: promote at the fp=0 operating point? (sign-off folds
relevance_policy_hash into governance_policy_hash)
- #000049: promote at 48% partial closure, or wait for Phase 3
semantic candidate selector?
Bench-maxing methodology codified in CLAUDE.md is the transferable
artifact: 'clean candidate-bench can mis-predict in BOTH directions
— real-data fixtures on both precision AND recall axes are the
only load-bearing measurement'. Eight instances across the two
arcs; the discipline applies to any future model-based addition.
Indexed in docs/TICKETS.md 'Distinction from other docs' section as
a non-ticket reference doc. Production verifier unchanged; nothing
in audit_mode; all work SHADOW pending sign-off.
Ran diagnose_deflection and diagnose_coherence on the 20+20 NEG
fixtures from step 3 parts A+B, plus all 808 STRICT.
Result: 0/40 NEG fire on diagnose_deflection (because the fixtures
are token-overlap-correct by construction — the question's subject
appears in the answer; that's exactly the failure mode #000052 was
built to catch beyond lexical). 0/40 on diagnose_coherence (the
fixtures are well-formed sentences).
So:
UNION (relevance OR deflection) = relevance alone (no lift)
INTERSECT (relevance AND deflection) = 0/40
Multi-signal combination doesn't help on these failure shapes.
But the architectural finding is positive in a different way: the
sidecars and the relevance reranker cover NON-OVERLAPPING failure
shapes cleanly:
- diagnose_coherence: structural breakage (word-salad, vacuous,
phrase-component-reuse). Owns the 'incoherent answer' slice.
- diagnose_deflection: token-overlap mismatch (subject anchor not
in answer). Owns the 'wholesale topic drift' slice.
- bge-large relevance: semantic aboutness mismatch despite shared
tokens. Owns the 'topic-collision / mis-cite / on-topic-but-
not-answering' slice — what §3.2 was built for.
Each signal owns its own slice; combining is redundant on these
cases. That's the architectural validation of the §3.1 + §3.2 +
existing-lexical-sidecars split as orthogonal, not overlapping.
Manifest runtime_viability.as_multi_signal_factor updated from
'viable' to 'TESTED — does not lift; sidecars are complementary
not combinatorial on these shapes'. Cleaned up stale
nli-shadow-grid-n1-minilm.json.
Also cleaned up a stale nli-shadow-grid JSON.
The §3.2 arc is now complete:
step 1 candidate-bench (overclaim, contrived data)
step 2 real STRICT FP (over-pessimistic 'NOT VIABLE')
step 3A real-context deflection (positive reversal, 55% at fp=0)
step 3B real-context mis-cite (100% at fp=0, motivating failure covered)
step 3C multi-signal combination (no lift; clean architectural split)
The recommended operating point holds: bge-reranker-large + cleaned
+ θ ≤ -2.42 → 100% mis-cite, 55% deflect, 0% STRICT FP.
Built bench/fixtures/5f/relevance-miscite-realcontext-v1.jsonl —
20 hand-crafted (claim, source) mis-cite pairs: claim about X, source
about Y, X≠Y but shared tokens. Each pair survives the lexical
title-relevance + verifier sidecars by construction. This is the
Zionist-entity failure mode (claim about a different entity than the
cited source, both lexically related).
Examples:
- 'Mercury is the smallest planet' / Roman-god Mercury source
- 'Java is a programming language' / Java-the-island source
- 'Apple Inc. was co-founded by Steve Jobs' / apple-the-fruit source
- 'The Eiffel Tower is in Paris' / Gustave-Eiffel-person source
- 'Mozart composed The Magic Flute' / Mozart-effect-theory source
Headline (bge-reranker-large + cleaned + θ ≤ -2.42):
Mis-cite catch: 20/20 = 100% ← FULL COVERAGE of motivating shape
Deflection catch: 11/20 = 55%
Combined NEG: 31/40 = 78%
Real-STRICT FP: 0/808 = 0% ← strictly safe
bge-large mis-cite scores: -9.37 to -3.81 (max). STRICT min: -2.42.
Mis-cite is STRICTLY SEPARABLE from real STRICT — there's a 1.4-pt
gap with no overlap. (Deflection harder; some overlap with weak STRICT.)
MiniLM-L-6 cost-pick (5× smaller, cleaned, θ=+3):
100% mis-cite + 65% deflect + 0.4% STRICT FP
Mis-cite is structurally MUCH easier than deflection — both models
hit 95-100% mis-cite catch at modest θ; deflection is harder because
'answer doesn't quite address question' can look like a weak STRICT.
That's appropriate: mis-cite is 'wrong topic entirely'; deflection
is 'right topic, not answering'.
Manifest:
- runtime_viability flipped (step 2 → step 3): NOT VIABLE → VIABLE
at the bge-large fp=0 operating point.
- demote_below_score still null pending fox+dav1d sign-off (setting
it folds relevance_policy_hash into governance_policy_hash per
#000049 §7 #2).
- PRIMARY RECOMMENDATION: bge-reranker-large + cleaned +
θ ≤ -2.42 → 100% mis-cite, 55% deflect, 0% STRICT FP.
Production verifier unchanged; still SHADOW. The §3.2 arc went:
candidate-bench (overclaim) → real STRICT step 2 (over-pessimistic
NOT VIABLE) → real-context deflection step 3 (positive reversal,
55% at fp=0) → real-context mis-cite step 3 part B (full closure
at fp=0 for the motivating failure mode). Eight meta-lesson
instances over the §000049 + §000052 arc, with the sharpest one
yet: clean candidate-bench can mis-predict in BOTH directions —
real-data fixtures on BOTH precision and recall axes are the
only load-bearing measurement.
Built bench/fixtures/5f/relevance-deflection-realcontext-v1.jsonl —
20 hand-crafted deflection answers against real bench-qa questions
(each: coherent well-formed answer using context-tokens, but NOT
addressing the question — the kind of failure §3.2 was built for).
Real-context deflection scores (cleaned, MiniLM-L-6):
min -6.71, median +1.56, max +6.34
Real-context deflection scores (cleaned, bge-reranker-large):
min -5.57, median -3.02, max +5.01
vs STRICT (cleaned):
MiniLM-L-6: min -5.29, p10 +6.90, median +9.32
bge-large: min -2.42, p10 +3.56, median +6.48
Distributions are CLEAN-separable on real data — 0% of MiniLM
deflections score above p10(STRICT); bge-large median is -3.02 vs
STRICT median +6.48.
Pareto frontier (MiniLM-L-6, cleaned):
θ=-5.29: catch 10% / fp 0% (fp=0 floor, low signal but safe)
θ=+3: catch 65% / fp 0.4% ← strong runtime soft-veto
θ=+5: catch 90% / fp 1.9% ← aggressive soft-veto
θ=+7: catch 100% / fp 10.4% (too FP for runtime)
bge-reranker-large:
θ=-2.42: catch 55% / fp 0% ← VIABLE runtime soft-veto at strict fp=0
The step-2 'NOT VIABLE' verdict was an artifact of using the
candidate-bench NEG distribution (tight contrived band, overlapped
real STRICT) as the recall denominator. Real-context deflections sit
in a much lower band (median around -3 for bge-large) than real
STRICT, so absolute thresholds DO separate them cleanly. The
relevance reranker IS a viable runtime soft-veto on this design —
the candidate-bench-only step 2 measurement misled us.
Manifest's runtime_viability.as_runtime_demotion_veto flipped from
'NOT VIABLE' to 'VIABLE at low-to-moderate FP', with both per-model
Pareto frontiers recorded. demote_below_score stays null pending
fox+dav1d sign-off; setting it folds relevance_policy_hash into
governance_policy_hash per #000049 §7 #2 discipline.
Still SHADOW; production verifier unchanged; no audit_mode effect.
Eighth instance of the meta-lesson, with the lesson sharpening AGAIN:
clean candidate-bench can MIS-PREDICT in BOTH directions — over-
optimistic on threshold (step 2) AND over-pessimistic on viability
(step 3 reversal). Real-data fixtures are the only load-bearing
denominators on either axis.
Tested Q→A score vs A→context_lead score on 808 pooled STRICT pairs
with cleaned MiniLM-L-6. The hypothesis was: deflected answers
should have HIGH A→ctx_lead but LOW Q→A (still grounded but
off-topic). Reality:
Q→A quantiles: p10 +6.90 p50 +9.32 p90 +10.57
A→ctx quantiles: p10 -6.73 p50 +1.51 p90 +8.68
Δ=Q→A-Actx: p10 +0.90 p50 +6.95 p90 +14.62
Of the 73 cleaned STRICT-fires at cb-θ=6.754:
- 71 (97%) have BOTH axes low (co-varying — not deflection)
- 2 (3%) match the deflection signature; both are the same
'who wrote GNU linux?' replica
Why it doesn't separate: the A→ctx axis isn't measuring what the
hypothesis assumed. A focused-claim against a 30KB topic-broad
Wikipedia haystack scores LOW by default — that IS the normal
STRICT shape (the answer is one clause in a sprawling document).
The cross-encoder expects the document to be ABOUT the query
(MS-MARCO retrieval shape); a STRICT (answer, full-haystack) pair
violates that. So both axes co-vary and the delta is a poor
discriminator.
This closes the third precision-side rescue path I'd left open
after step 2's verdict:
✗ absolute threshold — universal walk-back across 6 models
✗ cleaning preprocessing — helps but doesn't separate
✗ contrastive Q→A vs A→ctx — doesn't separate either
The relevance reranker conclusively cannot be a runtime
demotion-only veto on this design. The soft-signal uses
(render-tail, multi-signal advisory, etc.) remain viable.
runtime_viability.as_contrastive_signal updated from 'worth
measuring' to 'tested, doesn't separate'. Production verifier
unchanged; no audit_mode effect.
Hand-inspection of the bottom-15 STRICT-fires from the raw §3.2.2 step 2
sweep showed claim-lattice overlay markup ([E\d+ | title | hash: '…'])
depressing scores on correct concise answers (the 6× Henry-VIII case),
while true-positive deflections (broad-question / narrow-answer like
'winners of all major sports?' → just-one-sport) remained correctly
low-scored. So the noise FP class is the bracket metadata; cleaning it
should reduce FP without losing true-positive signal.
Built clean_for_relevance() in arborist/qa/relevance/shadow.py — strips
[E\d+ | ... ] blocks + trailing '...']' tails. Baked into
ShadowRelevance.check_question_answer / check_claim_source by default
(opt out with clean_input=False). relevance_shadow_sweep.py applies it
to inputs before _score_batch (opt out with --no-clean).
Re-ran the full 6-model sweep on the 808-cell pooled STRICT with
cleaning:
bge-reranker-large 21.8% → 18.6% (-3.2)
MiniLM-L-4-v2 15.0% → 9.4% (-5.6 pts, -37% rel)
MiniLM-L-6-v2 11.6% → 9.0% (-2.6)
MiniLM-L-12-v2 11.0% → 8.0% (-3.0)
bge-reranker-base 9.5% → 9.0% (-0.5)
MiniLM-L-2-v2 4.5% → 1.5% (-3.0 pts, -67% rel)
Universal improvement, every model better. Big surprise: MiniLM-L-2-v2
— the model that FAILED the candidate-bench separability (margin
-1.97, declared 'capacity floor') — has the LOWEST real-traffic FP
rate at its own cb θ (1.5%). Because L-2's compressed score range
gives it a low cb θ which few real STRICT pairs score below.
SEVENTH instance of 'candidate-bench doesn't predict real-traffic'.
Runtime-veto verdict UNCHANGED — still not viable; smallest fp=0 θ on
real STRICT is below the cb NEG max for every model, so at any
runtime-safe θ the catch on cb NEG is 0/12. But cleaning is now FREE
improvement for any soft-signal / advisory / contrastive use of the
relevance score. Hand-inspected bottom-10 post-cleaning confirms true-
positive deflection signal preserved.
All 6 swept rerankers (bge-large/base, MiniLM-L-2/L-4/L-6/L-12)
false-fire on 4.5–21.8% of real bench-qa STRICT at their
candidate-bench fp=0 θ. The smallest θ that yields fp=0 on real
STRICT is BELOW the candidate-bench NEG max for every model —
meaning at the runtime-safe θ, catch on the 12 candidate-bench NEG
= 0/12 across the board.
Structural reason: real bench-qa STRICT answers have a much wider
score distribution (bge-large STRICT: min -2.20, p10 +2.99, p50
+5.80, p90 +7.54) than the tight contrived candidate-bench POS
band. The bottom 10% of legitimate STRICT score below where the
candidate-bench NEG cases sat. Distributions overlap heavily; no
threshold separates them.
This is the §3.2 mirror of #000049 §7 #27's recall-side walk-back —
clean candidate-bench → fails real-pipeline gate. Same diagnosis:
lexical-candidate selection + cross-encoder scoring + hard threshold
doesn't survive real-pipeline heterogeneity.
Verdict: relevance reranker CANNOT be promoted to a runtime
demotion-only veto on this design. demote_below_score stays null;
manifest gains runtime_viability block documenting the negative
result + the still-viable advisory soft-signal uses (render-tail,
multi-signal advisory, contrastive Q→A vs claim→source delta).
bench/scripts/relevance_shadow_sweep.py + bench/results/
relevance-shadow-sweep-pooled808.json committed. Production
verifier unchanged; advisory only; no audit_mode effect.
